analysis
KSA
9/28/2021
Aalborg West
1 iteration
## function clustering:
## Cluster sizes: [4, 48, 13, 2, 25]
## r (mean): [0.42683 0.33065 0.23284 0.85964 0.29923]
## r (std): [0.38758 0.21144 0.19902 0. 0.20686]
## p (mean): [0.27966 0.07972 0.1685 0. 0.09244]
## r (weighted avg of means): 0.32397
##
## IDEC clustering:
## Cluster sizes: [15, 35, 6, 15, 21]
## r (mean): [0.44183 0.45664 0.63186 0.42152 0.24097]
## r (std): [0.21427 0.20911 0.17064 0.21457 0.19284]
## p (mean): [0.02537 0.02723 0. 0.05517 0.1358 ]
## r (weighted avg of means): 0.4107
##
## IDEC: {'iter': 350, 'acc': 0.47826, 'nmi': 0.18412, 'ari': 0.15315, 'L': 0.02144, 'Lc': 0.10083, 'Lr': 0.01135}5 iterations
## function clustering:
## Cluster sizes: [4, 48, 13, 2, 25]
## r (mean): [0.42683 0.33065 0.23284 0.85964 0.29923]
## r (std): [0.38758 0.21144 0.19902 0. 0.20686]
## p (mean): [0.27966 0.07972 0.1685 0. 0.09244]
## r (weighted avg of means): 0.32397
##
## IDEC clustering:
## Cluster sizes: [35, 12, 22, 10, 13]
## r (mean): [0.44077 0.43811 0.349 0.49337 0.40114]
## r (std): [0.21661 0.22642 0.21301 0.20071 0.23143]
## p (mean): [0.03548 0.05228 0.05457 0.02305 0.08443]
## r (weighted avg of means): 0.4186
##
## IDEC: {'iter': 50, 'acc': 0.47826, 'nmi': 0.20247, 'ari': 0.17069, 'L': 0.02165, 'Lc': 0.05743, 'Lr': 0.0159}Plot ASV’s with poor prediction
Max. 60% zeros
## function clustering:
## Cluster sizes: [4, 47, 13, 2, 24]
## r (mean): [0.42683 0.33394 0.23284 0.85964 0.30421]
## r (std): [0.38758 0.2127 0.19902 0. 0.20638]
## p (mean): [0.27966 0.0808 0.1685 0. 0.09071]
## r (weighted avg of means): 0.32722
##
## IDEC clustering:
## Cluster sizes: [22, 4, 25, 29, 10]
## r (mean): [0.33274 0.65596 0.34411 0.46175 0.49856]
## r (std): [0.23196 0.09897 0.2127 0.21464 0.19663]
## p (mean): [0.10345 0. 0.06707 0.02871 0.00545]
## r (weighted avg of means): 0.41026
##
## IDEC: {'iter': 50, 'acc': 0.36667, 'nmi': 0.13862, 'ari': 0.05385, 'L': 0.01457, 'Lc': 0.05048, 'Lr': 0.00952}Aalborg East
1 iteration
## function clustering:
## Cluster sizes: [7, 47, 10, 1, 17]
## r (mean): [0.27128 0.28037 0.2873 1. 0.25347]
## r (std): [0.20666 0.20823 0.22719 0. 0.19983]
## p (mean): [0.07096 0.11861 0.11314 0. 0.16499]
## r (weighted avg of means): 0.28364
##
## IDEC clustering:
## Cluster sizes: [31, 27, 16, 7, 1]
## r (mean): [0.26488 0.4806 0.47451 0.55528 1. ]
## r (std): [0.19733 0.20592 0.217 0.19528 0. ]
## p (mean): [0.11326 0.0209 0.02899 0.00032 0. ]
## r (weighted avg of means): 0.41057
##
## IDEC: {'iter': 100, 'acc': 0.36585, 'nmi': 0.06125, 'ari': 0.00837, 'L': 0.02855, 'Lc': 0.06645, 'Lr': 0.02191}5 iterations
## function clustering:
## Cluster sizes: [7, 47, 10, 1, 17]
## r (mean): [0.27128 0.28037 0.2873 1. 0.25347]
## r (std): [0.20666 0.20823 0.22719 0. 0.19983]
## p (mean): [0.07096 0.11861 0.11314 0. 0.16499]
## r (weighted avg of means): 0.28364
##
## IDEC clustering:
## Cluster sizes: [14, 21, 8, 22, 17]
## r (mean): [0.5155 0.57918 0.50708 0.30583 0.42396]
## r (std): [0.19827 0.17044 0.22934 0.20928 0.20838]
## p (mean): [0.0099 0.00146 0.0144 0.08399 0.03683]
## r (weighted avg of means): 0.45576
##
## IDEC: {'iter': 500, 'acc': 0.30488, 'nmi': 0.07715, 'ari': 0.00849, 'L': 0.01711, 'Lc': 0.10372, 'Lr': 0.00674}Plot ASV’s with poor prediction
Max. 60% zeros
## function clustering:
## Cluster sizes: [7, 46, 8, 1, 17]
## r (mean): [0.27128 0.28184 0.34645 1. 0.25347]
## r (std): [0.20666 0.20875 0.2416 0. 0.19983]
## p (mean): [0.07096 0.11856 0.07797 0. 0.16499]
## r (weighted avg of means): 0.29043
##
## IDEC clustering:
## Cluster sizes: [25, 18, 14, 15, 7]
## r (mean): [0.31481 0.57157 0.46224 0.53723 0.55528]
## r (std): [0.21394 0.17189 0.20063 0.18948 0.19528]
## p (mean): [0.08373 0.00053 0.02066 0.00638 0.00032]
## r (weighted avg of means): 0.46298
##
## IDEC: {'iter': 150, 'acc': 0.3038, 'nmi': 0.11311, 'ari': 0.01721, 'L': 0.01611, 'Lc': 0.08337, 'Lr': 0.00777}abund cluster 3
abund cluster 3
Randers
1 iteration
## function clustering:
## Cluster sizes: [10, 30, 3, 1, 14]
## r (mean): [0.35942 0.26386 0.35035 1. 0.24628]
## r (std): [0.25535 0.20577 0.15288 0. 0.21967]
## p (mean): [8.7630e-02 1.1810e-01 1.3000e-04 0.0000e+00 1.8291e-01]
## r (weighted avg of means): 0.29326
##
## IDEC clustering:
## Cluster sizes: [12, 1, 32, 1, 12]
## r (mean): [0.36814 1. 0.30185 1. 0.28074]
## r (std): [0.24432 0. 0.22608 0. 0.21303]
## p (mean): [0.06274 0. 0.10849 0. 0.09683]
## r (weighted avg of means): 0.33527
##
## IDEC: {'iter': 150, 'acc': 0.43103, 'nmi': 0.13993, 'ari': 0.00243, 'L': 0.04978, 'Lc': 0.04865, 'Lr': 0.04491}Notice only 1 ASV in cluster 1 and 3
IDEC cluster 0
IDEC cluster 1
IDEC cluster 2
IDEC cluster 3
What’s wrong with the y-axis scale for cluster 1 and 3?
Time series
5 iterations
## function clustering:
## Cluster sizes: [10, 30, 3, 1, 14]
## r (mean): [0.35942 0.26386 0.35035 1. 0.24628]
## r (std): [0.25535 0.20577 0.15288 0. 0.21967]
## p (mean): [8.7630e-02 1.1810e-01 1.3000e-04 0.0000e+00 1.8291e-01]
## r (weighted avg of means): 0.29326
##
## IDEC clustering:
## Cluster sizes: [16, 1, 1, 29, 11]
## r (mean): [0.25136 1. 1. 0.28136 0.74268]
## r (std): [0.18797 0. 0. 0.20949 0.12108]
## p (mean): [0.1208 0. 0. 0.11592 0. ]
## r (weighted avg of means): 0.38536
##
## IDEC: {'iter': 50, 'acc': 0.43103, 'nmi': 0.10639, 'ari': 0.05123, 'L': 0.01612, 'Lc': 0.04971, 'Lr': 0.01115}IDEC cluster 1
IDEC cluster 3
max 60% zeros
## function clustering:
## Cluster sizes: [10, 29, 3, 1, 13]
## r (mean): [0.35942 0.2702 0.35035 1. 0.27849]
## r (std): [0.25535 0.20762 0.15288 0. 0.21972]
## p (mean): [0.08763 0.11544 0.00013 0. 0.11901]
## r (weighted avg of means): 0.30538
##
## IDEC clustering:
## Cluster sizes: [22, 16, 6, 1, 11]
## r (mean): [0.33344 0.23949 0.59266 1. 0.74268]
## r (std): [0.21813 0.1643 0.24546 0. 0.12108]
## p (mean): [0.09326 0.11998 0.00175 0. 0. ]
## r (weighted avg of means): 0.42666
##
## IDEC: {'iter': 50, 'acc': 0.30357, 'nmi': 0.07736, 'ari': -0.05267, 'L': 0.01688, 'Lc': 0.05394, 'Lr': 0.01149}Notice only 1 ASV in IDEC cluster 3
IDEC cluster 3
time series ASV75
Randers max 60% pseudo-zeros (0.01%)
## function clustering:
## Cluster sizes: [8, 26, 2, 1, 11]
## r (mean): [0.37068 0.29845 0.56577 1. 0.29391]
## r (std): [0.22799 0.21134 0. 0. 0.23404]
## p (mean): [0.08934 0.09429 0. 0. 0.11815]
## r (weighted avg of means): 0.3352
##
## IDEC clustering:
## Cluster sizes: [6, 18, 8, 1, 15]
## r (mean): [0.4675 0.26281 0.45508 1. 0.45134]
## r (std): [0.24751 0.18712 0.27757 0. 0.22612]
## p (mean): [0.00273 0.12641 0.04351 0. 0.02956]
## r (weighted avg of means): 0.39471
##
## IDEC: {'iter': 50, 'acc': 0.45833, 'nmi': 0.224, 'ari': 0.04011, 'L': 0.01316, 'Lc': 0.03514, 'Lr': 0.00964}Notice only 1 ASV in IDEC cluster 3 
Aalborg West max 60% pseudo-zeros (0.01%)
## function clustering:
## Cluster sizes: [4, 45, 10, 2, 24]
## r (mean): [0.42683 0.34519 0.26916 0.85964 0.30421]
## r (std): [0.38758 0.21462 0.19932 0. 0.20638]
## p (mean): [0.27966 0.07623 0.10634 0. 0.09071]
## r (weighted avg of means): 0.34062
##
## IDEC clustering:
## Cluster sizes: [16, 18, 20, 12, 19]
## r (mean): [0.45088 0.53866 0.3636 0.40843 0.39661]
## r (std): [0.20315 0.18684 0.22379 0.22408 0.21179]
## p (mean): [0.02786 0.01358 0.06789 0.04816 0.03302]
## r (weighted avg of means): 0.43081
##
## IDEC: {'iter': 300, 'acc': 0.37647, 'nmi': 0.16606, 'ari': 0.07004, 'L': 0.02644, 'Lc': 0.11045, 'Lr': 0.01539}Aalborg East max 60% pseudo-zeros (0.01%)
## function clustering:
## Cluster sizes: [4, 43, 7, 1, 15]
## r (mean): [0.2217 0.29408 0.35684 1. 0.25882]
## r (std): [0.17115 0.21392 0.25573 0. 0.20627]
## p (mean): [0.08213 0.11292 0.08864 0. 0.17057]
## r (weighted avg of means): 0.29875
##
## IDEC clustering:
## Cluster sizes: [6, 10, 20, 18, 16]
## r (mean): [0.47692 0.56524 0.59479 0.31753 0.43775]
## r (std): [0.17282 0.18849 0.16184 0.22182 0.20421]
## p (mean): [0.00021 0.00015 0.00036 0.08444 0.02129]
## r (weighted avg of means): 0.47328
##
## IDEC: {'iter': 50, 'acc': 0.3, 'nmi': 0.08632, 'ari': -0.00566, 'L': 0.02166, 'Lc': 0.07334, 'Lr': 0.01433}Aalborg West, allow ASVs not in func guilds for IDEC
## function clustering:
## Cluster sizes: [4, 45, 10, 2, 24]
## r (mean): [0.42683 0.34519 0.26916 0.85964 0.30421]
## r (std): [0.38758 0.21462 0.19932 0. 0.20638]
## p (mean): [0.27966 0.07623 0.10634 0. 0.09071]
## r (weighted avg of means): 0.34062
##
## IDEC clustering:
## Cluster sizes: [148, 152, 182, 124, 137]
## r (mean): [0.33734 0.41272 0.3905 0.3688 0.32149]
## r (std): [0.20517 0.20695 0.21077 0.22365 0.19798]
## p (mean): [0.07464 0.04121 0.04944 0.06921 0.07549]
## r (weighted avg of means): 0.36811
##
## IDEC: {'iter': 2050, 'acc': 0.23149, 'nmi': 0.03288, 'ari': -0.00284, 'L': 0.05158, 'Lc': 0.11829, 'Lr': 0.03975}Avedøre
## function clustering:
## Cluster sizes: [5, 2, 8, 19, 25]
## r (mean): [0.38072 0.65749 0.66608 0.3619 0.28188]
## r (std): [0.1992 0. 0.21081 0.23707 0.20675]
## p (mean): [0.10079 0. 0.0178 0.12202 0.18316]
## r (weighted avg of means): 0.38085
##
## IDEC clustering:
## Cluster sizes: [146, 107, 140, 129, 111]
## r (mean): [0.36379 0.5596 0.57512 0.47394 0.49837]
## r (std): [0.22615 0.21949 0.22519 0.22284 0.21978]
## p (mean): [0.12466 0.03468 0.02445 0.05319 0.0467 ]
## r (weighted avg of means): 0.48968
##
## IDEC: {'iter': 1600, 'acc': 0.24329, 'nmi': 0.04148, 'ari': 0.00247, 'L': 0.02401, 'Lc': 0.14228, 'Lr': 0.00978}Damhusåen A
## function clustering:
## Cluster sizes: [4, 5, 11, 18, 32]
## r (mean): [0.22182 0.56418 0.3469 0.41021 0.37367]
## r (std): [0.17733 0.25791 0.24186 0.28031 0.26191]
## p (mean): [1.6075e-01 1.0000e-04 1.3014e-01 1.1729e-01 1.0061e-01]
## r (weighted avg of means): 0.38379
##
## IDEC clustering:
## Cluster sizes: [88, 130, 129, 104, 118]
## r (mean): [0.56829 0.43377 0.32222 0.55812 0.46991]
## r (std): [0.22037 0.23508 0.21312 0.20058 0.21861]
## p (mean): [0.02216 0.06111 0.10597 0.01419 0.03946]
## r (weighted avg of means): 0.45951
##
## IDEC: {'iter': 1550, 'acc': 0.24956, 'nmi': 0.03085, 'ari': 0.0075, 'L': 0.04109, 'Lc': 0.13676, 'Lr': 0.02741}Damhusåen B
## function clustering:
## Cluster sizes: [4, 5, 11, 25, 33]
## r (mean): [0.26839 0.47913 0.39725 0.3948 0.39974]
## r (std): [0.15492 0.22326 0.23854 0.2827 0.25805]
## p (mean): [0.13805 0.0216 0.07992 0.10807 0.08694]
## r (weighted avg of means): 0.39616
##
## IDEC clustering:
## Cluster sizes: [113, 117, 104, 102, 121]
## r (mean): [0.4595 0.41946 0.47897 0.33844 0.51827]
## r (std): [0.2323 0.23459 0.21518 0.21547 0.21717]
## p (mean): [0.04857 0.06546 0.03128 0.09816 0.02306]
## r (weighted avg of means): 0.44532
##
## IDEC: {'iter': 1350, 'acc': 0.25135, 'nmi': 0.04636, 'ari': 0.00586, 'L': 0.03187, 'Lc': 0.12696, 'Lr': 0.01918}Damhusåen C (errored)
Damhusåen D
## function clustering:
## Cluster sizes: [4, 5, 12, 27, 33]
## r (mean): [0.2603 0.52826 0.55769 0.38368 0.31656]
## r (std): [0.21316 0.32165 0.22179 0.27173 0.25321]
## p (mean): [0.24976 0.09116 0.02083 0.09505 0.14507]
## r (weighted avg of means): 0.38494
##
## IDEC clustering:
## Cluster sizes: [139, 126, 97, 145, 130]
## r (mean): [0.38789 0.45211 0.49249 0.47677 0.57268]
## r (std): [0.22818 0.24132 0.23108 0.23036 0.21045]
## p (mean): [0.07787 0.06156 0.03866 0.04024 0.01568]
## r (weighted avg of means): 0.47447
##
## IDEC: {'iter': 1050, 'acc': 0.2449, 'nmi': 0.0449, 'ari': 0.00685, 'L': 0.03344, 'Lc': 0.13902, 'Lr': 0.01953}Hirtshals
## function clustering:
## Cluster sizes: [1, 2, 11, 31, 29]
## r (mean): [1. 0.63527 0.37099 0.33957 0.46462]
## r (std): [0. 0. 0.24731 0.22927 0.2725 ]
## p (mean): [0. 0. 0.13375 0.13499 0.06909]
## r (weighted avg of means): 0.41016
##
## IDEC clustering:
## Cluster sizes: [134, 158, 105, 130, 85]
## r (mean): [0.49488 0.65588 0.3881 0.40792 0.40206]
## r (std): [0.20385 0.18592 0.21917 0.23584 0.23509]
## p (mean): [0.03194 0.00653 0.08606 0.08912 0.09714]
## r (weighted avg of means): 0.48676
##
## IDEC: {'iter': 1950, 'acc': 0.25327, 'nmi': 0.03346, 'ari': -0.00172, 'L': 0.03558, 'Lc': 0.16297, 'Lr': 0.01928}Mariager
## function clustering:
## Cluster sizes: [2, 4, 13, 24, 32]
## r (mean): [0.5144 0.35971 0.36289 0.26995 0.36809]
## r (std): [0. 0.21516 0.26593 0.1918 0.23886]
## p (mean): [0. 0.05995 0.1098 0.14155 0.10876]
## r (weighted avg of means): 0.33924
##
## IDEC clustering:
## Cluster sizes: [100, 191, 120, 147, 167]
## r (mean): [0.39569 0.33861 0.50986 0.32232 0.44601]
## r (std): [0.21486 0.23142 0.2321 0.21964 0.21777]
## p (mean): [0.06721 0.11577 0.03598 0.11978 0.04771]
## r (weighted avg of means): 0.39626
##
## IDEC: {'iter': 2200, 'acc': 0.28, 'nmi': 0.02822, 'ari': 0.00121, 'L': 0.05278, 'Lc': 0.14793, 'Lr': 0.03799}Randers, again
## function clustering:
## Cluster sizes: [8, 1, 9, 5, 25]
## r (mean): [0.36888 1. 0.46214 0.24662 0.30174]
## r (std): [0.23392 0. 0.25093 0.16545 0.21275]
## p (mean): [0.09857 0. 0.02277 0.06693 0.09053]
## r (weighted avg of means): 0.35181
##
## IDEC clustering:
## Cluster sizes: [133, 140, 127, 135, 90]
## r (mean): [0.39161 0.36953 0.4898 0.42262 0.3901 ]
## r (std): [0.24162 0.23132 0.22163 0.22105 0.21586]
## p (mean): [0.06984 0.07522 0.02733 0.04231 0.05824]
## r (weighted avg of means): 0.4131
##
## IDEC: {'iter': 1400, 'acc': 0.248, 'nmi': 0.03712, 'ari': 0.00287, 'L': 0.03344, 'Lc': 0.12776, 'Lr': 0.02066}